Fashion passes; style remains. (quote attributed to Gabrielle “Coco” Chanel)
Style may be elusive to achieve but thanks to big data analytics, fashion selfies and “trendspotters”, fashion trends can be predicted. As with any type of forecasting or sticking a finger in the air (if we are to be glib), there can be bad calls. Bad calls can mean millions of dollars worth of unsold goods – cue J. Crew and Gap’s recent troubles. But it can be done and companies like WGSN, Editd, Trendstop, etc. sell their trend forecasting services to retail companies for hundreds of thousands of dollars; predicting what consumers are willing to buy a year from now, season after season.
Well, the field could get even more interesting if we can get an algorithm to automate what’s hot and what’s not. Computer scientists from Spain recently presented at the Computer Vision and Pattern Recognition conference in Boston, MA outlining an algorithm that analyzes the fashionability of an outfit in a picture and makes recommendations on how to improve it. Fashionability, as defined in the paper (here in PDF) came from analyzing 144,169 posts from fashion community chictopia.com. It is obviously still in concept stage but it is an interesting concept to pursue especially if you can match it with image recognition such as the one developed by Google and Stanford. I mention Google’s image recognition technology efforts in particular because its main business is still search. I don’t know about you but being able to have intelligent search results for what’s on trend (whether in fashion, home, technology) right now, the next season or the next year sounds like a tantalizing future. But if you already have style – that undefinable, born-with-it, unteachable, elusive thing where you just know – then maybe that future is already here for you.
On a personal note, I received some great feedback on my post about buying a house here and I’m glad I can help. I will most likely share more on that journey especially on what I learned about renovating and share more of the tools I use to make life easier so stay tuned!
Here are this week’s most relevant news in tech and retail:
- FCC to fine AT&T $100M for throttling Internet connections on unlimited wireless data plans
- Password manager LastPass hacked; users are urged to change master password and to enable 2-factor authentication
- California rules that Uber drivers are employees and not contractors; could push up costs for ride-sharing startup
- Tesla Motors borrows $750M from series of banks including Bank of America, JP Morgan Chase, Deutsche Bank
- Fitness tracking company, Fitbit shares surge 48.4% in IPO debut, IPO price at $20
- CVS buys Target’s in-store pharmacies for $1.9B; deal expected to close this year
- US cosmetics group Coty to buy some of P&G’s beauty brands for $12B
- Gap Inc. to close 175 of Gap stores over the next few years and layoff 250 employees from headquarters
- Online luxury retail website Net-a-Porter reports record sales in final set of full-year results (paywall)
- Online handmade marketplace Etsy launches fund to crowdfund new products; makes Wall Street happy
3 thoughts on “Automating “style” and the weekly roundup in tech and retail”
Fascinating! These are the kinds of topics that no one else is covering. Thank you, Leah Grace! Excellent info.
Very interesting ! Thank you!